I'm Moved!

I have moved this blog. If your browser does not automatically redirect, please click here.

Change of Software?

Now that I have gone over 50 posts, it looks like I will keep this blog going (thanks to the loyal reader out there!). Right now, the blog is hosted by blogger. I have now installed wordpress on my own machine, and have transferred things over. Having my own wordpress installation allows me more flexibility in add-ons for the blog. Can anyone make the case for keeping it on blogger? You can compare at the new system.

Optimize magazine and Operational Excellence

Optimize is a 70,000 circulation magazine aimed at CIOs and CTOs. The topic of the May, 2006 issue is "Operational Excellence", and it contains a host of articles about operations research, one done by yours truly. My article, entitled "CIO as Business Predictor" tries to talk about the role uncertainty plays in decision making and how OR approaches can help address these issues. The sidebars in the article describe the finalists in this year's Edelman Competition.

A second article, entitled "Bringing Unity to Cardinal Health", described how IT and operational excellence go hand-in-hand. There is a sidebar of an interview with Irv Lustig from ILOG on the how IT and OR interact.

The Editor's Note to the issue also talks about OR quite a bit, and says some very nice things about me! I think I will get it framed and send to my mother.

Between this and the coverage by Stephen Baker from Business Week, OR is getting noticed quite a bit these days in the more mainstream press.

Stephen Baker on Operations Research

Stephen Baker, the author of the Business Week cover story on how Math Will Rock Your World, is rapidly becoming a highly visible writer about Operations Research. His May 4 blog entry, entitled "Why journalists don't cover how things work" had some comments on his experience at the INFORMS Practice Conference:

At the O.R. conference (the association is called INFORMS), there were far too many interesting presentations for one person to cover them all. The people behind operations at Intel, IBM, the Army, Ford and plenty of others provided inside looks. Beat reporters of those companies could have feasted on these lectures. But they weren't there.

Why? The press covers news, stocks, companies and personalities. But try pitching a cover story on operations. People think it's ... boring. Trouble is, if we want to know where things are going, we have to understand how they work. And when the process is transformative, as it often is in OR, there's nothing boring about it. The winner of the annual Informs Franz Edelman award, by the way, was the Warner Robins Air Logistics Center. They overhauled the maintenance of jumbo C-5 transport aircraft, reducing repair time by 33%. This means that these monsters, which cost taxpayers $2.3 billion each, spend more time in the air and less time in the shop.

If Stephen keeps this up, we may see lots more press coverage.

Great Newsletter

The publishers of OR/MS Today (the membership magazine of INFORMS) have started an online and email newsletter. The first issue contains a large number of short, interesting tidbits from the world of (practical) OR. Feel free to subscribe so it shows up in your mailbox as soon as it is published.

Edelman Competition

This year's Edelman Competition will take place in a few days at the INFORMS Practice Conference. The Edelman Competition presents projects in practical operations research that have a significant impact on an organization. This year's finalists include work from Travelocity, the TSA, Warner Robbins Air Logistics, Center and more. I really like the Edelman: the finalists put papers into Interfaces, and those papers are a tremendous resource for my MBA classes.

Of course, every competition could be improved. I was mulling on the thought of turning this competition into a reality show. Lock a bunch of ORers in a plant for six weeks with cameras everywhere, and watch them try to out-optimize each other to be the last one standing. Probably wouldn't compete with American Idol, but I certainly would watch it!

IBM's Center for Business Optimization

I am in New York to give a talk to IBM's Center for Business Optimization. Bill Pulleyblank is heading this activity. Bill has had a really interesting career: he started in academia, and did really fundamental work in combinatorial optimization. He then moved to IBM, starting at Watson Research and moving on to doing things like heading the Blue Gene Project, which created the world's fastest supercomputer. CBO is a startup with IBM that tries to merge the assets of IBM (software packages, etc.) with the consulting skills of IBM Business Consultants (formerly PriceWaterhouseCooper's consulting arm).

It is exciting to see a company like IBM take optimization seriously. The projects I have chatted with people about look like "real" optimization and business analytics: data mining and modeling approaches to fraud dectection (both tax fraud and Medicare fraud), supply chain optimization, marketing design, and so on. They have a number of case studies that outline their various projects.

Operations Research Job Prospects

Money Magazine and salary.com have a ranking of 166 job titles, based on salary and job prospects. I was happy to see "College Professor" as the second best job (good salary, good growth, lots of freedom). Having Operations Research Analyst mired in roughly 120th place (out of 166) was less fun to see. The salary for the field is good, but job growth was relatively low. Still, OR Analyst beat out mathematician, economist, physicist, librarian and many other seemingly appealing fields.

Descriptive versus Prescriptive

Working in a business school (the Tepper School at Carnegie Mellon), many of my colleagues are economists (or "financial economists" as many of my finance colleagues are titled). One of the big hurdles we have in communicating is a differing view of the purpose of models. For many economists, models are used to describe behavior. For instance, a model of certain types of incentives may lead to particular outcome behavior. If we see the outcome behavior, this suggests the model is a good one. If an economist does not see the outcome behavior, then there is a puzzle at best, and a bad model at worst.

For an ORer (a word I just made up, because OR person sound stilted), models are almost always prescriptive: they tell you what to do. For that same model, an ORer will, if happy with the model, not worry about whether the outcome behavior is happening. If not, then people are doing things wrong, and should smarten up and follow the OR prescription.

Articles like the one today in the New York Times (subscription required) make me feel happier about the OR approach. People, even reasonably sophisticated people, just don't seem to make good decisions. The example is drawn from finance, but examples abound:

MANY index funds track the Standard & Poor's 500, but they differ from one another in one major respect: their fees. You'd think that it would be obvious to investors to pick the fund that charges the least. But you'd be wrong.

In fact, this truth was anything but obvious to a group of elite students. In an elaborate simulation created by several researchers, many students at Harvard and the Wharton School of the University of Pennsylvania failed to select the lowest-cost index fund for their portfolios, even when they were all but spoon-fed the right answer.

...

In the first simulation, the professors asked 30 undergraduates at Harvard and Wharton, as well as 83 M.B.A. students at Wharton, to allocate a hypothetical $10,000 among four S.& P. 500 index funds that would be held for one year. All four funds invested in the same 500 companies, and matched the index's allocation for each stock, so the only significant difference in the funds' returns would come from their fees — their front-end loads, or sales charges, and their management expenses. And because of the way the professors designed the experiment, these four funds had relatively high fees, ranging from 3.09 percent to 5.89 percent.

Each student received copies of the four index funds' prospectuses, which varied from 26 to 116 pages long. The prospectuses provided detailed descriptions of the funds' loads and expenses, along with a great deal of other information. To encourage the students to take the exercise seriously, each was paid a small amount — $5 for undergraduates, $20 for M.B.A. students — and was told that one student at each school, picked at random, would receive any profit that the $10,000 portfolio produced over the 12 months.

The rational response, the professors argue, would have been to allocate all the money to the fund with the lowest fees. Yet fewer than 20 percent of either group of students did so. As a result, the hypothetical portfolios built by most of the students paid much higher fees than were necessary: 1.22 percentage points more, on average, among the undergraduates and 1.12 points higher among the M.B.A. students.

The work was done by James J. Choi, an assistant professor of finance at the Yale School of Management; David Laibson, an economics professor at Harvard; and Brigitte C. Madrian, a professor of business and public policy at the Wharton School. A copy is at http://www.som.yale.edu/faculty/jjc83/fees.pdf.

Makes me happy to be prescriptive!

New MLB Scheduler

Some of you may know that I have been working with Major League Baseball for a number of years on scheduling issues. Finally, in 2005 they played a schedule created by a small firm that I am a partner in. Unfortunately, we got beaten out for the 2006 schedule, as MLB.com reports. We'll get them in 2007!

Want to learn more about Constraint Programming?

Thom Frühwirth and Slim Abdennadher have a book called "Essentials of Constraint Programming". Best of all, Thom has more than 400 powerpoint slides on his website to supplement the book.

David Simchi-Levi Presentation

David Simchi-Levi is here at CMU today speaking on inventory systems where the decision maker is not risk neutral. David is a professor at MIT and Editor-in-Chief of Operations Research. He also runs his own company, LogicTools. I think he has given up sleep.

It is surprising that there is so little research on risk-averse decision makers in the operations management context. This clearly is a useful integration of economic theory and operations research and also involves the interface between finance and operations in a firm. In David's work, even marketing comes into play since prices can be set in the model in order to influence demand.

Work like this often spends time in pages of greek letters with subscripts, and this paper is no different (though David is an excellent lecturer, so it was not as mind-numbing as some lectures like this).

With fixed costs for ordering, even models with risk neutral decision makers are hard to solve. Normally, the optimal decisions are set with an (s,S,p) policy: if inventory is under s, order up to S, and set price p. But suppose the inventory level is a bit above s. Should price be set high in order to decrease demand or should the order go up to S, and price be set low. This example shows that the price has a discontinuity, making it hard to get characteristics of the solution.

With risk aversion, things get more complicated. For the case where there is no fixed ordering cost, the optimal policy is a base stock policy, but the base stock level depends on the wealth level of the decision maker. With fixed costs, for exponential utility, the optimal inventory policy is independent of wealth, making things a bit easier (though it is still hard to figure out the exact policy).

One interesting aspect of this that came up in Q&A is the need to combine pricing and inventory decisions. Most companies are not aligned this way: marketing sets price, while operations sets inventory. Even in these complicated models, though, the results generally look like "Operations, do (s,S); Marketing, set price p". This suggests very tight integration is not needed: coordination is enough.

When you add financial hedging aspects, there is still this decoupling aspect: the financial decisions do not affect the operational decisions.

It would take someone more skilled than I to explain these results to a general audience, but I find it interesting that models that contain all of marketing (price setting), operations (inventory setting) and finance (hedging) are amenable to analytical solution. This seems a very rich research area.

A Stonewall Connection to Operational Research

My parents grew up on farms outside a then-small (now medium) sized town in Manitoba named Stonewall. For a period in the early 1900s, a boy named Charles Goodeve lived in Stonewall. He lived there for about 10 years, before his family moved to Winnipeg. There is an article in the Stonewall Argus (the Gordon Trick mentioned is my father) about his life, including his work during World War II in the British Navy. Among other things, he figured out how to protect ships from underwater mines through a demagnetization process.

The operational research connection? In 1948, Sir Charles Goodeve founded the OR Club, which would later become the Operational Research Society.

Tournament Time!

The NCAA Tournament is irresistable to OR types. Predicting the tournament has proven a rich area of application. Jay Coleman of the University of North Florida has a scorecard approach that gives probabilities of wins for every game in the first round. For three of the 32 games, his approach favors the lower seeded team (including number 10 Alabama over number 7 Marquette. As far as number 1 seeds go, his method gives a 99%+ chance to Villanova and Duke in their first game (a number 16 has never beat a number 1) but just 97% for UConn and 95% for Memphis.

Joel Sokol of Georgia Tech has done a lot of work in this area. He has some interesting comments on the 2006 brackets.

INFORMS has some other pointers in this area.

Anyone else like to talk about their OR approaches to this?

Operations Research and CIOs

United Airlines now has a CIO who is also responsible for operations research and other activities. This seems a natural, if somewhat unusual move (OR is often under manufacturing, operations or some other structure). OR is all about using information, and as firms realize the value of information (and CIOs) is in their ability to extract knowledge from information, more OR may be in the hands of the CIO.

Applied Mathematical Programming

One of my all time favorite textbooks is Applied Mathematical Programming by Bradley, Hax, and Magnanti. This text is now available on the web!

Passing of Joan Wingo

The field of operations research is full of unsung heros: people who make the field better by doing their jobs with enthusiasm, creativity, and skill. If you have published in Operations Research over the last six years, or if you have read and admired papers in that journal, you will have seen the work of Joan Wingo, Managing Editor of Operations Research. Joan worked with Editor-in-Chief Larry Wein to make Operations Research run, and run well. I worked with Joan over the last few months on transition issues with the new editorial board.

Joan, sadly and untimely, passed away February 13. Her obituary notice contains some thoughts of her friends, coworkers and loved ones. We, as a field, owe her a great deal.

New Editors at Operations Research

The INFORMS journal Operations Research has a new editorial board, led by Editor-in-Chief David Simchi-Levi. David asked me to take on the area of OR Forum. I was hesitant to take on an editorial duty (handling papers has turned out to be an Achilles heel of mine), but the opportunity to handle papers that inspire discussion and controversy was too much to turn down. Here is the area statement:

The OR Forum area invites work that challenges the reader to consider and evaluate the status of past, present, or future prospects and challenges within the field of operations research. Possible submissions include critical reviews of research in a specialized field, closely reasoned commentary on the practice within an area, analysis of prospects for operations research broadly, or any other area where a substantive, significant work will clarify and illuminate research and practice. Published work will often be accompanied by supplemental pieces that enhance or dispute the theses developed.


An online forum will provide opportunity to continue the discussion after publication. Papers that address prospects in areas not traditionally covered by Operations Research are strongly encouraged, as are provocative papers that take a strong stand on policy and practice issues. The arguments made in the paper should not be casual or speculative, but should be based on a firm foundation consistent with publication in a professional journal. Survey papers are appropriate providing such papers go beyond a listing of who wrote what to include a critical appraisal of the research and the prospects for the future. The work should be accessible and of interest to a significant portion of the
readership of Operations Research. Authors are encouraged to contact he Area Editor early in the process of developing their work to determine suitability for consideration in this area.

Any thoughts for suitable articles?

PhD salaries

Forbes magazine has an article about the lack of mathematically trained US workers. Most of the article is about outsourcing, but the issu of starting salaries came up:

A person fresh from graduate school with a Ph.D. in operations research can make $90,000 at SAS Institute--far less than the $150,000-plus salaries top MBAs can command. "Yes, fine, we need to pay more," Steve Odland, CEO of Office Depot admitted.

Well, kinda... First, a fresh MBA rarely makes $150,000: the average Tepper MBA graduate is closer to $90,000. Second, the lifestyle of those who make the high amounts in terms of stress, travel, and so on is pretty rotten, at least by my standards. Finally, there are lots of fresh Ph.D.s in OR making more than $90,000, often teaching those same MBAs!

Not to say that Ph.D.s should make lots of money (I am one of them myself!), but that doesn't seem the most dire aspect: it is really the lack of supply of interesting jobs that allow the true use of an OR Ph.D.

Doing research already done?

The New York Times has a nice article about how research is often rediscovered. The lead begins about operations research:

In 1996, Rakesh Vohra, a professor at Northwestern University, and his colleague Dean Foster published "A Randomized Rule for Selecting Forecasts," a paper in the journal Operations Research. It illustrated how a random investor could outperform a group of professional stock pickers simply by following a "buy and hold" investment strategy.
Skip to next paragraph
Alain Pilon

It was important research, the authors believed, until they learned that the same discovery had been made at least 16 times since the 1950's. And no one, Dr. Vohra said, ever realized they were not doing original work.

As a referee, there are certain things in sports scheduling that I get quite often (generally some variation of de Werra's work on minimum break scheduling) and I recently went a long way on a paper before de Werra pointed out to me that the results were included in a somewhat more obscure publication of his. I wonder if online search will make it easier to find these duplicate results before it makes the literature?

Happenings in Cork

I have just returned from a site visit to the Cork Constraint Computation Centre (4C: it's a pun, don't you know) in Cork Ireland, where I am on the advisory board. Ireland is a country that seems to have its government investments set up well. A few years ago, they a identified Gene Freuder as someone they would like to attract and made him a very nice offer. Gene is one of the (or perhaps the main) founders in the area of constraint programming, and he has put together an extremely impressive lab. Lots of activity, both in fundamental research and in industrial outreach. The lab isn't perfect (more OR!), but it is very, very good. If you are learning more about constraint programming, this lab's site is a good place to start.

Additional Online info at Business Week

There is an additional article online (only) associated with the "Math will Rock Your World" article in Business Week. Entitled Search Advertising by the Numbers, the article concentrates on the problem of bidding on keywords. Here is a quote:

Khan, an accountant by training, looks at the process like a linear programming problem. That may sound exotic to mathophobes, but the concept is simple. He sets a goal that he aims to maximize. In Khan's case, that's revenue. Then he factors in the constraints, such as cost per keyword.

This can get complicated. If he bids for the word "loan," for example, he competes with much of the lending industry. That drives up the price. What's more, it's a big, broad word, which means the results are much more complex than a far cheaper targeted keyword phrase such as "Tuscaloosa mortgage."

"If we can get one dollar out of an auto loan, we can get $20 if we do a mortgage loan," Khan says. "A keyword like 'loan' can deliver either of the two products. So should we buy the keyword based on potential of auto loan or mortgage loan?" He has his team do the math

As always, the article doesn't go into any real detail of models or algorithms, but it does make our field much more visible.

Business Week

Just as I was bemoaning the lack of press coverage, Business Week comes up with a cover story entitled "Math Will Rock Your World". The field of operations research is even mentioned by name, which is a tremendous step for the area!

Does your library have all the INFORMS Journals?

INFORMS is running a grass-roots effort to get more libraries to carry INFORMS journals. They have a neat setup that you can check if your (university) library carries every journal. You can even email your librarian to encourage subscriptions.

Most libraries are cutting back on journals, partially due to some extremely high prices on journals. INFORMS journals, being published by INFORMS, are quite reasonably priced, but of very high quality. Every library should have them!

Sad News from India

From the Indian Express:

Terror struck an international conference at the Indian Institute of Science (IISc) campus on Wednesday night killing a retired Mathematics professor from the Indian Institute of Technology, Delhi, M.C. Puri, and seriously injuring four others including one of the inventors of the Simputer, Prof Vijay Chandru from IISc.

The attack was, according to eyewitness accounts, carried out by a lone gunman who wielded an AK-47 and threw hand grenades. The attacker was driven away in a white Ambassador car immediately after the attack, eyewitnesses said.

All major cities in south India have been put on high alert after the incident.

The other injured persons have been identified as Dr Pankaj Gupta from Delhi, P Patel, a lab assistant at the Cadila Lab in the IISc campus, and a woman identified only as Sonia, an assistant professor at IIM, Lucknow.

‘‘All the injured have been ruled to be out of danger,’’ Additional Commissioner of Police H C Kishore Chandra said.

Delegates at the International Conference on Operations Research Applications in Infrastructure Development and the 38th Annual convention of Operation Research Society of India (ORSI) were proceeding from IISc’s National Science Seminar Complex to the Satish Dhawan auditorium for an AGM of the ORSI at around 7:30 p.m. when the attack took place.

Full Story

Vijay Chandru is a co-author of mine, and my thoughts are with him, the others injured and the family of Prof. Puri.